ORIGINAL RESEARCH published: 04 March 2021 doi: 10.3389/fpls.2021.637214

Polyploidy on Islands: Its Emergence and Importance for Diversification

Heidi M. Meudt 1*, Dirk C. Albach 2, Andrew J. Tanentzap 3, Javier Igea 3, Sophie C. Newmarch 4, Angela J. Brandt 5, William G. Lee 5 and Jennifer A. Tate 4*

1 Museum of New Zealand Te Papa Tongarewa, Wellington, New Zealand, 2 Institute of Biology and Environmental Sciences, University of Oldenburg, Oldenburg, Germany, 3 Ecosystems and Global Change Group, Department of Plant Sciences, University of Cambridge, Cambridge, United Kingdom, 4 School of Fundamental Sciences, Massey University, Palmerston North, New Zealand, 5 Manaaki Whenua – Landcare Research, Dunedin, New Zealand

Edited by: Whole genome duplication or polyploidy is widespread among floras globally, but Natascha D. Wagner, traditionally has been thought to have played a minor role in the evolution of island University of Göttingen, biodiversity, based on the low proportion of polyploid taxa present. We investigate five Germany island systems (Juan Fernández, Galápagos, Canary Islands, Hawaiian Islands, and Reviewed by: Daniel Crawford, New Zealand) to test whether polyploidy (i) enhances or hinders diversification on islands University of Kansas, and (ii) is an intrinsic feature of a lineage or an attribute that emerges in island environments. United States Marc S. Appelhans, These island systems are diverse in their origins, geographic and latitudinal distributions, Georg-August-University levels of plant species endemism (37% in the Galapagos to 88% in the Hawaiian Islands), Goettingen, Germany and ploidy levels, and taken together are representative of islands more generally. *Correspondence: We compiled data for vascular plants and summarized information for each genus on Heidi M. Meudt [email protected] each island system, including the total number of species (native and endemic), generic Jennifer A. Tate endemicity, chromosome numbers, genome size, and ploidy levels. Dated phylogenies [email protected] were used to infer lineage age, number of colonization events, and change in ploidy level

Specialty section: relative to the non-island sister lineage. Using phylogenetic path analysis, we then tested This article was submitted to how the diversification of endemic lineages varied with the direct and indirect effects of Plant Systematics and Evolution, a section of the journal polyploidy (presence of polyploidy, time on island, polyploidization near colonization, Frontiers in Plant Science colonizer pool size) and other lineage traits not associated with polyploidy (time on island, Received: 03 December 2020 colonizer pool size, repeat colonization). Diploid and tetraploid were the most common Accepted: 11 February 2021 ploidy levels across all islands, with the highest ploidy levels (>8x) recorded for the Canary Published: 04 March 2021 Islands (12x) and New Zealand (20x). Overall, we found that endemic diversification of Citation: Meudt HM, Albach DC, our focal island floras was shaped by polyploidy in many cases and certainly others still Tanentzap AJ, Igea J, Newmarch SC, to be detected considering the lack of data in many lineages. Polyploid speciation on the Brandt AJ, Lee WG and Tate JA (2021) Polyploidy on Islands: islands was enhanced by a larger source of potential congeneric colonists and a change Its Emergence and Importance in ploidy level compared to overseas sister taxa. for Diversification. Front. Plant Sci. 12:637214. Keywords: colonization, diversification, endemism, island floras, ploidy level, phylogenetic path analysis, doi: 10.3389/fpls.2021.637214 polyploidy, whole genome duplication

Frontiers in Plant Science | www.frontiersin.org 1 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

INTRODUCTION phylogenetic context was lacking in previous studies, which is important because the phylogeny will indicate the number Since the time of Darwin (1839), the study of island biotas of origins on the island as well as patterns of diversification has provided major advances that have profoundly influenced on the island related to polyploidy or other factors. ecological, evolutionary, and biogeographical theory (MacArthur As robust (and increasingly, dated) phylogenetic hypotheses and Wilson, 1967; Mayr, 1967; Carlquist, 1974; Emerson, 2002). have amassed over the last several decades, along with continued Island biotas are generally the net outcome of immigration efforts to document chromosome numbers, we now have the (dispersal and establishment), local diversification, and extinction capability to test the role of polyploidy in contributing to (Carlquist, 1974), and these processes are known to be influenced species diversification on islands in a phylogenetic context by specifics of the island, such as age, area, distance from (Kellogg, 2016; Crawford and Archibald, 2017). As species in nearest potential source floras Rosenzweig,( 1995), and local many different island floras are known to have diverse habitat (abiotic and biotic) conditions (Savolainen et al., 2006; chromosome numbers and form species-rich groups, they may Vamosi et al., 2018). Islands may also pose selective filters not be chromosomally static lineages as previously thought that may be apparent in the intrinsic traits of species, including (Soltis et al., 2009; New Zealand: Murray and de Lange, 2011; those increasing their ability to disperse, colonize, and establish Canary Islands: Caujapé-Castells et al., 2017). A recent analysis in novel habitats (Crawford et al., 2009; Vargas et al., 2015). of the global distribution of polyploids indicated polyploid Among those traits, whole genome duplication or polyploidy frequency is highest in temperate areas rich in perennial herbs, has been suggested to be central to facilitating long-distance including mountainous areas (Rice et al., 2019). Analyzed from dispersal (Linder and Barker, 2014), the survival of small a global perspective, several island systems are polyploid-rich, populations (Rodriguez, 1996), and evolution of novel traits including the Hawaiian Islands (50% of analyzed species are (Soltis et al., 2014), all features of many island floras. Polyploids polyploid), New Zealand (46%), and Galápagos Islands (46%), are broadly categorized as autopolyploids, when formed within whereas others have lower polyploid frequencies – such as a species, or allopolyploids, when formed between genetically the Canary Islands (32%) – or insufficient data (Juan Fernández: distinct species (Stebbins, 1947). Polyploidy, especially 0 of 2 species included were polyploid; Rice et al., 2019). allopolyploidy, may also be a mechanism for increasing the However, that study was focused primarily on assessing external genetic diversity of the colonizer (Crawford and Stuessy, 1997; drivers of polyploid plant distribution globally, not on island Carr, 1998), which is often low for island colonizers and lineages systems, and did not utilize dated phylogenies. A dated (e.g., Frankham, 1997; but see García-Verdugo et al., 2015). phylogenetic context allows a more precise determination of Thus, polyploidy could have multiple advantages, particularly when diversification and polyploidization have occurred within in island floras. each lineage (i.e., as separate events), and puts the focus on Despite its potential benefits, polyploidy has been (multiple) lineage ages rather than a single island age. Such historically suggested to play a minor role in diversification an approach can generate multiple independent data points of island floras, with many groups showing “chromosomal that can be analyzed together to address the timing and role stasis” (Carr, 1998; Stuessy and Crawford, 1998). For the of polyploidy on islands. Dated phylogenies are especially oceanic island systems that inform this perspective (Hawaiian, important in the context of island diversification because they Juan Fernández, Galápagos, and Bonin Islands), make it possible to estimate in situ diversification rates and paleopolyploidy was suggested to have helped some lineages to consider in the analyses the geological processes (e.g., volcano establish, with little polyploidization thereafter Carr,( 1998; eruptions, inundations) that may affect lineages at different Stuessy and Crawford, 1998), such as in Gossypium (Malvaceae; evolutionary time scales. Figure 1). Thus, while polyploid, these oceanic island lineages We hypothesize that polyploidy has played an important remained chromosomally static. By contrast, chromosomal role in the diversification of island floras by facilitating dispersal variation was found in two island systems near their continental and establishment of plants to islands, and/or by generating source, the Queen Charlotte Islands and Canary Islands additional diversity through varying ploidy levels. To test a (Stuessy and Crawford, 1998). Although Crawford et al. conceptual model for how polyploidy influences species (2009) began to address the question of the evolution of diversification on islands, we synthesized published chromosome polyploidy in island systems and its role in diversification and divergence time data for 150 lineages representing 1,805 (for Asteraceae; Crawford et al., 2009), others discounted endemic species across the Juan Fernández, Canary, Hawaiian, its impact (Stuessy et al., 2014; Crawford and Archibald, and New Zealand archipelagos (Figure 1). All these island 2017). However, in those studies (Carr, 1998; Stuessy and systems, except New Zealand, have been included in previous Crawford, 1998; Crawford et al., 2009), chromosome counts comparative studies or reviews of polyploidy and diversification from only a small percentage of native island species were (see above), but without the time-calibrated phylogenetic context available, so interpretations were based on island origin we add here. By using phylogenetic path analysis, (continental vs. oceanic) and island age rather than polyploidy we simultaneously tested the strength and direction of causal itself. Other island systems, in particular New Zealand, have associations to explain the diversity of endemic island species not been included in these larger comparative studies, despite in these four archipelagos. We predicted that island lineages chromosome numbers being widely available for many native would be more diverse (i.e., have more endemic species) if plant species (see Dawson, 2000, 2008). Furthermore, a they (Figure 2): (P1) contained multiple ploidy levels; (P2)

Frontiers in Plant Science | www.frontiersin.org 2 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

ABCD E

FIGURE 1 | Outline and representative polyploid plants from the five analyzed island systems. From left to right:(A) New Zealand: Veronica topiaria (Plantaginaceae) © Phil Garnock-Jones; (B) Canary Islands: Aeonium arboreum (Crassulaceae) © Dirk Albach; (C) Hawaiian Islands: Argyroxiphium sandwicense (Asteraceae) © Marc Appelhans; (D) Juan Fernández Islands: Eryngium bupleuroides (Apiaceae) © Lukas Mekis; (E) Galápagos Islands: Gossypium darwinii (Malvaceae) © A. Emmerson.

repeatedly colonized islands with different ancestral species, indicated by a lack of monophyly. Hypotheses P1–P4 test effects of ploidy that are both direct (P1) or indirect (P2–P4) and mediated by time (P2), changes to the ploidy levels themselves (P3), or source pool size (P4). Alternatively, hypotheses P1, P2, and P5 test polyploidy as a trait important in diversification, whereas P3, P4, P6, and P7 test polyploidy as a trait important in dispersal or establishment.

MATERIALS AND METHODS Data Collection FIGURE 2 | Conceptual model for how species diversity in island lineages We compiled data of indigenous species, subspecies, varieties, varies with polyploidy and extrinsic colonization history. Arrows show and forms (hereafter: species), and noted whether they were predicted linkages between direct and indirect effects of polyploidy (P1–P4, native or endemic, for vascular plants on five island systems shown in black) and non-polyploidy effects (P5–P7, shown in gray) on diversification of island floras. Left-hand of the plot shows traits intrinsic to using published references for the Canary Islands (CI; island lineages whereas the right-hand shows traits associated with island Arechavaleta et al., 2010), Galápagos Islands (GI: Jaramillo dispersal and colonization. Díaz et al., 2017, 2018), Hawaiian Islands (HI: Wagner et al., 2005 and electronic updates; Imada, 2012), Juan Fernández Islands (JF: Table 5.1 in Stuessy et al., 2017), and New Zealand had more time to generate ploidy levels once on these islands, (NZ: Breitwieser et al., 2012; Schönberger et al., 2019). We then as indicated by the length of time they were present on an added data for each species (where known and when from a island; (P3) had a different (i.e., higher) ploidy level relative sample from that island system) on chromosome number (Index to their sister lineage, indicating polyploidization immediately to Plant Chromosome Numbers,1 Chromosome Counts Database,2 before or after colonization of the island; (P4) were derived and additional literature search), genome size (Pellicer and from a large pool of overseas congeners and thus the likelihood Leitch, 20203) and ploidy level. We summarized this information of polyploidization events that could enhance diversity was for each genus on each island system, including the total greater. We also tested other direct explanations that did not number of indigenous and endemic species, whether the genus involve ploidy and predicted island lineages would be more was endemic to the island system, chromosome number, genome diverse if they (Figure 2); (P5) were older, because they have size, and ploidy level (Table 1). For each genus with at least simply had more time to undergo speciation and isolate one native species on a particular island group, we identified populations across a greater availability of niches (e.g., Lee et al., 2012; Tanentzap et al., 2015); (P6) were derived from 1http://legacy.tropicos.org/Project/IPCN a large pool of overseas congeners and thus the probability 2http://ccdb.tau.ac.il/ of colonization from that pool would be higher; and (P7) 3https://cvalues.science.kew.org/

Frontiers in Plant Science | www.frontiersin.org 3 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

TABLE 1 | Summary of island systems for all genera and species1 of native vascular plants.

Archipelago

Variable New Zealanda Canary Islandsb Hawaiian Islandsc Juan Fernándezd Galápagose

No. of islands 3 main islands plus 7 main islands plus 6 8 main islands, plus 129 3 islands 18 main islands, ~600 smaller islands/ smaller islands/islets smaller islands/islets plus 110 smaller islets islands/islets/rocks Area of islands 268,021 km2 7,493 km2 28,311 km2 100.2 km2 7,985 km2 Shortest distance from nearest 1,490 km E of Australia 96 km W of Morocco 3,500 km SW of 580 km W of Chile 930 km W of continental neighbor United States mainland Ecuador Main reference for data in rows Schönberger et al. Arechavaleta et al. (2010) Wagner et al. (2005 and Stuessy et al. (2017) Jaramillo Díaz below (2019)2 electronic updates); et al. (2017, 2018) Imada (2012) No. native species 2,551 1,6773 1,233 209 550 No. endemic species (%) 2,096 (82%) 647 (39%) 1,082 (88%) 136 (65%) 204 (37%) No. species with chromosome 1,962 (77%) 1,171 (70%) 414 (34%) 87 (42%) 39 (7%) numbers (%) No. species with genome size 245 (12.5%) 237 (14%) 29 (2%) 0 0 estimates (%) No. native genera 430 466 272 103 277 No. endemic genera (%) 48 (11%)4 22 (5%) 30 (11%) 12 (11%) 7 (3%) No. genera with multiple ploidies on 88 (21%/35%) 96 (21%/40%) 27 (10%/19%) 5 (5%/13%) 3 (1%/3%) island (% of total/% of genera with at least two native species) Range of ploidy levels in one genus 1–6 1–4 1–2 1–2 1–2 Most common ploidy levels (range) 2x, 4x, 6x (2x–20x) 2x, 4x (2x–12x) 2x, 4x (4x–6x) 2x, 4x (2x–8x) 2x, 4x (2x–4x) No. genera with two or more native 256 (60%) 239 (51%) 139 (51%) 39 (38%) 105 (38%) species on island (%) No. genera5 for which a molecular 205 (48%) 170 (36%) 86 (32%) 31 (30%) 71 (26%) phylogeny is available (% of native genera) No. genera5 for which a dated 112 (26%) 41 (9%) 27 (10%) 8 (8%) 4 (1%) molecular phylogeny is available (% of native genera) No. total genera (%) / no. genera5 (%) 389 (92%)/250 (98%) 406 (87%)/219 (92%) 155 (57%)/92 (67%) 58 (56%)/25 (64%) 21 (8%)/13 (12%) for which there is at least one known chromosome number6 No. species per genus range/ 1–143/6.0/2 1–33/3.6/2 1–80/4.5/2 1–12/2.0/1 1–19/2.0/1 average/median No. gymnosperm species (% 22 (100%) 6 (33%) 0 0 0 endemic) No. gymnosperm genera (% endemic) 10 (30%) 3 (0%) 0 0 0 No. fern species (% endemic) 210 (46%) 50 (6%) 167 (75%) 57 (46%) 125 (9%) No. fern genera (% endemic) 57 (4%) 23 (0%) 51 (4%) 24 (2%) 52 (0%)

Island geographic data from: aMcGlone et al. (2001). bCarracedo and Troll (2016). cPrice (2004). dStuessy et al. (1998). eRivas-Torres et al. (2018). 1Includes species, subspecies, varieties, and forms; collectively called “species” for simplicity. 2Most data generally agree with Breitwieser et al. (2012). 3The species numbers here include 288 subspecies, which means they will be higher than in Arechavaleta et al. (2010) which only counts species. 4But see Garnock-Jones (2014) who suggested the actual number of endemic genera is lower, ca. 28–44 (7–10%). 5With two or more native species on the island. 6From a specimen of at least one native species from the island. the most recent phylogenetic study (that included at least estimated using WebPlotDigitizer4 on dated tree figures. We also one native species sampled from that island) to determine estimated the variance for each crown and stem age using the if the genus was monophyletic on the island system and standard deviation (SD) or assuming normality of the longer whether there was a change in ploidy level compared to the of the two tails of the 95% highest posterior density or confidence sister group. For lineages with dated phylogenies, we extracted intervals (after Lee et al., 2012; Tanentzap et al., 2015; the mean time of their divergence from both sister lineages Brandt et al., 2016). For genera that were not monophyletic and their most recent common ancestor (i.e., stem and crown ages, respectively). Values were either given in papers or 4https://automeris.io/WebPlotDigitizer/

Frontiers in Plant Science | www.frontiersin.org 4 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands within the archipelago, we used dates of the earliest-derived were highly correlated with crown age (Pearson correlation species or lineage on the island system. From these data (see between mean ages, r = 0.84, df = 110, p < 0.001), and may Supplementary Files), we calculated several comparative better reflect the entire evolutionary history of clades Scholl( statistics (Table 2), as described below. and Wiens, 2016). The model of archipelago ploidy levels was identical to that for endemic diversity except without including Statistical Analyses ploidy levels and archipelago monophyly as predictors and letting We tested our conceptual model (i.e., P1–P7) by fitting two the stem age effect vary with the island system. Following separate phylogenetic least squares (PGLS) regression models standard practice, the models were fitted assuming the expected using the gls function (implemented in the nlme package; covariance in the responses between any two lineages was Pinheiro et al., 2014) in R v.3.6, which were united into a proportional to their shared evolutionary history along a single path analysis with the piecewiseSEM package (Lefcheck, phylogenetic tree, i.e., a Brownian motion (Symonds and Blomberg, 2016). As we were interested in the effect of ploidy on 2014). Distances were derived by pruning the largest time- diversification, we analyzed lineages that had at least two native calibrated phylogenetic tree available for vascular plants, which species on the island system, a dated phylogeny available with contained 74,531 species and was generated with hierarchical at least one native island species, and at least one chromosome clustering analysis of GenBank data and a backbone provided count of a native island specimen (Supplementary File 1). by Open Tree of Life version 9.1 (Smith and Brown, 2018; Jin Lineages that fit our criteria were included from New Zealand and Qian, 2019). On average, sister branches in this phylogeny (n = 98), Canary Islands (n = 23), Hawaiian Islands (n = 23), had an overlap of 1,792 base pairs, which corresponds to roughly and the Juan Fernández Islands (n = 6), but data from Galápagos one or two gene regions. Although the responses were Islands were insufficient for downstream analysis. These lineages counts and so could also be modeled with other approaches were always genera aside from four exceptions that were closely (e.g., phylogenetic generalized linear models), log-normal related genera that formed a monophyletic group on the transformations made the responses normally distributed, as archipelago (Supplementary File 1). The models were fitted to expected for some Poisson distributed variables. Thegls function the number of endemic species and number of ploidy levels also had the advantage that it could be used to fit a path for each lineage on each island system. To predict species analysis and incorporate uncertainty in the responses unlike number, we modeled observations based on both log-transformed these other regression approaches. We specifically accounted stem age and number of ploidy levels of the island lineage, for uncertainty in divergence time estimates and different levels total number of accepted and unresolved species in the lineage of data completeness by weighting observations of species counts outside of the island system according to The Plant List v1.1 and ploidy levels with the inverse square root of divergence (2013) , and separate binary variables for whether the lineage time SDs and the proportion of species with chromosome was monophyletic on the island system and its ploidy level counts, respectively, afterGaramszegi and Møller (2010). was different from its closest sister lineage outside of the island system. We let the effects of ploidy levels vary among the four island systems and accounted for differences in the mean endemic RESULTS diversity of lineages among island systems. Doing so also ensured that island systems with more data points, such as New Zealand, Comparison of Island Groups did not bias the estimated effects towards themselves. We used The five island systems (New Zealand, Canary Islands, Hawaiian stem ages to represent lineage age because they were better Islands, Juan Fernández, and Galápagos Islands) differ regarding sampled than crown ages (n = 150 vs. 116 lineages, respectively), number and total area of island system, distance to nearest

TABLE 2 | Summary of variables used for the statistical analyses.

Island archipelago

Variable New Zealand Canary Islands Hawaiian Islands Juan Fernández All islands

No. of lineages 98 23 23 6 150 Number of monophyletic island lineages 33 13 18 3 67 Number of lineages with different ploidy level than closest sister 8 2 6 2 18 outside of island Mean (SD) number of island endemic species per lineage 11.3 (17.6) 13.7 (15.0) 16.4 (19.0) 4.5 (3.7) 12.2 (17.1) Mean (SD) stem age per lineage in millions of years 15.63 (17.30) 8.48 (6.19) 9.21 (6.23) 5.93 (5.86) 13.16 (14.81) Mean (SD) number of ploidy levels per lineage 1.7 (1.2) 1.5 (0.7) 1.2 (0.4) 1.0 (0.0) 1.6 (1.1) Mean (SD) % of endemic island species with chromosome counts 86.0 (22.8) 74.3 (24.0) 50.9 (31.9) 71.8 (36.1) 78.2 (27.8) Median number of species in lineage outside of island system 34 77 58 384 51

Means and standard deviations (SD) were calculated for each non-count variable. The data set included the subset of 150 lineages of native vascular plants with the following criteria: dated phylogeny available, at least one native species sampled in the phylogeny, at least two native species on the island system, and at least one known chromosome count from the island system.

Frontiers in Plant Science | www.frontiersin.org 5 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands continent, number (and percentage) of native and endemic levels represented on the archipelago, ranging from 2x to 20x. species and genera, and data availability for phylogenies, dated Of the 199 NZ genera with phylogenies, 40% have multiple phylogenies, chromosome numbers, genome size, and ploidy ploidies (the majority with two or three ploidy levels). information (Table 1). The island systems have a five-fold The Canary Islands rank second after NZ in terms of number difference in the number of indigenous genera (103 in JF to of native species and data availability, third in terms of area, 466 in CI), a 10-fold difference in the number of native species and first in terms of proximity to the nearest mainland Table 1( ). (209 in JF to 2,551 in NZ), and a 20-fold difference in the They have the second-lowest percentage of endemic species number of endemic species (136 in JF to 2,056 in NZ). For (39%). There are 27 genera with 10 or more species, including simplicity, we chose to lump sub-specific ranks under species, several adaptive radiations, two of which comprise larger lineages so our determination of absolute species numbers may differ of multiple genera, i.e., the Aeonium alliance (Crassulaceae; from other sources and we acknowledge that there is much 62 endemic species in four genera) and the Sonchus alliance taxonomic uncertainty in island radiations. We do not think (Asteraceae; 62 endemic species in six genera; Kim et al., 2008). that this approach has affected our analyses other than minor Gymnosperms are rare in CI (only six native species, two of differences in species numbers. which are endemic), and of the 50 fern species, only 6% are On all island systems, the majority of indigenous vascular endemic; there are no endemic genera of gymnosperms or plant species are angiosperms, with a small percentage of ferns. Of the 466 genera, 217 have between 2 and 33 species, ferns and even fewer gymnosperms. The average number of and the rest are monotypic on the archipelago. Of these 217, native species per genus is 2.0 (JF, GI), 3.6 (CI), 4.5 (HI), 78% have a molecular phylogeny but only one-third are dated and 6.0 (NZ). The percentage of endemic genera was low phylogenies. Between 1 and 27 native CI species per genus for all island systems (3–11%), as was the availability of (17–100%) are included in the phylogenies; over half have genome size data (up to 14% in CI). The availability of 50% or more species included in phylogenies. Roughly 20% chromosome numbers was highest for CI and NZ (70–77%) of these genera with phylogenies are monophyletic or nearly and lowest for GI (7%), with HI (34%) and JF (42%) also so in CI (i.e., one CI origin likely), 30% are not monophyletic having relatively few counts. The number of native in CI (i.e., more than one origin in CI), and about half are non-monotypic genera on each island system ranged from unknown due to lack of phylogeny or sampling. Eighty-seven 39 (38%, JF) to 256 (60%, NZ). On all island systems, molecular percent of the genera have at least one species with a chromosome phylogenies have been published for the majority of these count, and there are genome size estimates known for about non-monotypic genera, ranging from 62% in HI to 80% in 14% of species. Ninety-four percent of genera with phylogenies NZ, but dated phylogenies are less prevalent (4% in GI to have at least 50% of the species with chromosome counts. 44% in NZ). Genera have between 1 and 4 ploidy levels represented on Of all the island systems studied here, New Zealand has the islands, ranging from 2x to 12x, with the most common the largest area, the lowest number of main islands (but the ploidies being diploid and tetraploid. Of the 170 genera with highest number of total islands including smaller islands), the phylogenies, 40% have multiple ploidies (the majority with largest flora, and the highest percentage of data available for two ploidy levels). its species (Table 1). Ferns comprise about 9% of species (210 The Hawaiian Islands have the second-largest land mass species, 46% endemic) and 13% of genera in the native plant (after New Zealand) of the five island systems and the highest vascular flora [57 genera, of which only monotypicLeptolepia level of species endemicity (88%; Table 1). There are 272 native (Dennstaedtiaceae) and Loxsoma (Loxsomaceae) are endemic]. vascular genera, 139 (51%) have between 2 and 80 species, With respect to gymnosperms, all 22 species and one-third and the remainder are represented by one species on the of the 10 native genera are endemic. Of the 430 NZ vascular archipelago. Ferns represent 14% of the vascular plant species plant genera, 256 have at least two native species, and the diversity and show a high level of endemicity at the species rest are monotypic on the archipelago. Of these 256, 78% level (75%). There are no native gymnosperm taxa in HI. have a phylogeny but only 46% have dated phylogenies. About Phylogenies are available for about one-third of the genera, 70% of genera with a phylogeny (144 genera) have 50% or and of these, one-third are dated. Overall, we identified 142 more species included in phylogenies. Roughly one-third of phylogenies that included Hawaiian taxa, and 39% of these these genera are monophyletic or nearly so in NZ (i.e., one were for groups in which only one species occurs in HI. The NZ origin likely), one-third are not monophyletic in NZ (i.e., flora is not well characterized chromosomally at the species more than one origin in NZ), and about 40% are unknown level overall (34% of species with known chromosome numbers), due to lack of phylogeny or sampling. Ninety percent of the but there is broad representation at the generic level, with genera have at least one species with a chromosome count, 57% of genera having at least one species known. Ploidy levels but there are few published genome size estimates (12.5%, see range primarily from diploid to octoploid, with diploid and Table 1). Fifty-seven genera have 10 or more native species, tetraploid representing the majority of known levels. Overall, and two of these have over 100 species, i.e., Veronica polyploid series with multiple ploidy levels are lacking but (Plantaginaceae, 143 species, 96% endemic) and Carex there are true dysploid series (e.g., x = 13, 14) that occur in (Cyperaceae, 118 species, 88% endemic). Ninety-four percent several genera of the Hawaiian silversword alliance (Asteraceae, of genera with phylogenies have at least 50% of the species Carr, 1998). The most common scenario for HI taxa (with with chromosome counts. Genera have between 1 and 6 ploidy known chromosome numbers) is that these lineages are tetraploid

Frontiers in Plant Science | www.frontiersin.org 6 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands relative to their overseas congeners and these show chromosomal only 21 genera, and of these 13 genera have two or more stasis on the island system [e.g., angiosperms: the six genera species. Data for chromosome numbers for native taxa and of Campanulaceae, Bidens (Asteraceae), Stenogyne (Lamiaceae); dated phylogenies were only available for one lineage, which ferns: Deparia (Athyriaceae)]. Among angiosperm genera with was not included in the statistical analyses. Where chromosome two or more species, 60 (43%) are the result of a single numbers are available, the species are primarily 4x or 6x with colonization event and 11 are considered to have diversified just one ploidy level in those lineages in most cases. In two in situ. Representative adaptive radiations include the genera cases, multiple ploidy levels exist on the archipelago and these of the Lobelioideae (Campanulaceae), those in the Lamioideae were both in fern genera [Adiantum (Pteridaceae) and Polypodium (Lamiaceae), and the aforementioned silversword alliance (Polypodiaceae)]. (Asteraceae; Price and Wagner, 2018). Twenty-seven genera are the result of two colonization events and 10 have three Statistical Analysis of Island System Data or more introductions to HI. Table 2 summarizes the lineages from the four island systems The Juan Fernández Islands have the smallest flora and the that fitted the criteria for the statistical analyses (i.e., at least smallest area, have the same low number of only three main two native species on the island system, dated phylogeny islands as New Zealand, and are probably second to last in available with at least one native island species included, and terms of data availability (Table 1). Of the 103 JF genera, 39 at least one chromosome count of a native island specimen). have between 2 and 12 species, and the rest are monotypic New Zealand lineages comprised 66% of the data set, followed on the archipelago. Only three genera have more than 10 by the Hawaiian and Canary Islands with 15% each, and native species, i.e., Carex (Cyperaceae, 11 species including Juan Fernández at 3% (no data were available for the Galápagos seven endemic), Dendroseris (Asteraceae, 12 species, all endemic), Islands). Just under half (45%) of the genera were monophyletic and Hymenophyllum (Hymenophyllaceae, 11 species, only one on the island system. For the majority of lineages where the endemic). Overall, species endemicity is high (65%), placing ploidy level of the sister group could be determined, the JF as third among the island systems in terms of species-level lowest ploidy level on the island was the same as the sister endemicity, and equal with New Zealand and the Hawaiian group (JF 60% n = 6, HI 73% n = 23, CI 92% n = 23, NZ Islands (11%) at the generic-level. Ferns represent about 92% n = 98; Table 2). The remaining lineages had a one-quarter of native vascular plant species (57 species, 46% polyploidization event that occurred sometime along the stem endemic) and genera (24 genera, of which only monotypic lineage immediately before or after colonization of the island Thyrsopteris is endemic); there are no native gymnosperms. and thus may represent island neopolyploidization. New Zealand There are phylogenies for 33 genera but only eight of these had the oldest mean stem age (15.63 vs. 4.18 million years are dated. Between 1 and 7 native JF species (20–100%) are in Juan Fernández) and the highest number of ploidy levels included in the phylogenies; 21 genera (63% of those with a per lineage (1.7 vs. 1.0 in Juan Fernández). The Hawaiian phylogeny) have 50% or more species included in phylogenies. Islands had the highest mean number of island endemic species Roughly 20% of these genera are monophyletic or nearly so per lineage (16.1 vs. 4.8 in Juan Fernández) but also the in JF (i.e., one JF origin likely), 20% are not monophyletic lowest mean percentage of species with chromosome counts in JF (i.e., more than one origin in JF), and the remaining (50 vs. 86% in New Zealand). 60% are unknown due to lack of phylogeny or sampling. Only Our statistical analysis supported the hypothesis that about half of the JF genera have at least one species with a polyploidy shapes the diversification of island florasFigure 2 ( ). chromosome count, and there are no genome size estimates We found that greater levels of polyploidy directly promoted known. JF genera have either one or two ploidy levels represented endemic diversity on island systems (P1; t150 = 3.11, p = 0.002). on the archipelago, ranging from 2x to 8x, with diploids and Over the range of observed ploidy levels (1–6), the estimated tetraploids the most common. Of the 33 JF genera with number of species in island genera increased from a mean of phylogenies, about 70% have at least one species with a 1 to 4 [95% confidence interval (CI) for increase: 0.7–9.9]. chromosome count and half have at least 50% species with a Polyploidy itself was enhanced by a larger source of potential chromosome count. Of these, almost all have only one ploidy level. congeneric colonists (P4; t150 = 5.36, p < 0.001) and a change

The Galápagos Islands have the highest number of main in ploidy level from overseas sister taxa (P3; t150 = 4.04, islands with an area only slightly greater than the Canary p < 0.001). Lineages that changed in ploidy near the time of Islands (Table 1). Native vascular genera total 277, seven are island colonization had, on average, 4.4 ploidy levels as compared endemic (3%); 105 (38%) of these have between 2 and 19 with 2.5 levels where these changes were absent (95% CI for species, and the remaining are monotypic on the archipelago. difference: 0.83–3.22). In these same lineages, as the size of Ferns represent ca. 23% of vascular plant species (9% of these the potential colonist pool increased from the 25th to 75th are endemic); there are no endemic fern genera. Like the percentile of observed values (5–275 species in the source Hawaiian and Juan Fernández Islands, there are no native pool), estimated ploidy levels increased from a mean of 3.5 gymnosperms in the Galápagos. Molecular-based phylogenies to 5.4 (95% CI for increase: 1.0–2.9). are available for 42 genera, and only four of these are dated Lineage age also affected diversification outcomes. Older phylogenies. Between 1 and 13 GI species (11–87%) are included lineages were more diverse (t150 = 3.69, p < 0.001), as expected in the phylogenies; 21 genera have 50% or more species included if they had more time to diversify (P5) and had slightly more in the phylogenies. Chromosome numbers are available for ploidy levels across island systems (P2; t150 = 1.98, p = 0.049).

Frontiers in Plant Science | www.frontiersin.org 7 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

Over the interquartile range of observed stem ages, ploidy Previous authors have already commented that polyploidy levels were estimated to increase from a mean of 4.1 to 4.7 may assist colonization of islands by facilitating the establishment (95% CI for increase: 0.4–2.1). The effect of stem age on ploidy of species (Linder and Barker, 2014) and this also parallels levels was, however, reversed (negative and statistically significant) findings that polyploidy is more common in invasive plants (Te for the Hawaiian Islands (Figure 3, “island effect”), resulting Beest et al., 2012). Furthermore, polyploids seem to be preadapted in fewer ploidy levels in the older Hawaiian lineages (t150 = −2.60, to island colonization. One factor providing polyploids an p = 0.010). Endemic diversity was also higher, on average, for advantage over related diploids in the colonization of new

Hawaiian lineages (t150 = 2.11, p = 0.036) and lower in Juan habitats is higher genetic diversity, which leads to lower Fernández lineages (t150 = −2.10, p = 0.038; Figure 3, “island inbreeding depression of polyploids (Rosche et al., 2017) and effect”). Whether lineages were monophyletic on the island improved adaptability (Scarrow et al., 2020). Other factors system (P7) or had more potential congeneric colonists outside include improved environmental tolerance (Moura et al., 2020), of the island system (P6) had no direct effect on endemic higher frequency of vegetative reproduction, and selfing (in diversity (t150 = 1.62, p = 0.107 and t150 = 1.62, p = 0.108, part due to a breakdown of self-incompatibility; Robertson respectively). Tests of directed separation indicated a missing et al., 2011), which provide a means to colonize and survive path in our analysis from the lineages being monophyletic longer at low population size (Baker, 1967; Herben et al., 2017). (“repeat colonization”) to ploidy levels (“polyploidy,” Figure 2). Thus, there are several arguments for considering polyploids Including this path in our final model made no difference to to be superior colonizers but only very few that support the our results (t150 = 0.15, p = 0.885). Overall, the model predicted idea that polyploids are better dispersers (e.g., Kuo et al., 2016 both endemic diversity and the number of ploidy levels reasonably for ferns). For example, Harbaugh (2008) suggested that polyploid well (R2 = 0.35 and 0.38, respectively; Figure 3). Santalum (Santalaceae) have smaller seeds and fruits with a thicker endocarp, making them easier to disperse by birds. However, easy dispersal does not necessarily translate to frequent DISCUSSION establishment and consequently more species on an island. Our analyses demonstrate that polyploidy contributes Understanding evolutionary processes on islands has improved significantly to diversification of established island lineages. with molecular phylogenetic analyses and especially the Endemic richness on islands increases over time in colonizing development of time-calibrated phylogenies, which provide a lineages, reflecting the accumulation of new species through temporal framework for island colonization. Using this local speciation on islands. The results of the path analysis information from five island systems, we highlight the important indicate that polyploidy influences species diversity on islands role of polyploidy in the dispersal, colonization, and diversification through colonizing species belonging to large lineages, of island lineages. Our results overturn the perception that apparently able to speciate more than others independent chromosomal stasis is a feature of island systems, and instead of the geographic setting. Thus, the source pool size is one provide evidence that demonstrates the importance of polyploidy factor that drives polyploidization, which in turn affects in promoting both colonization and species diversification. endemicity on the archipelago. Additionally, ploidy levels also increase over time and thereby contribute to increasing the number of endemic taxa. According to our path analysis (Figure 3), lineages are more likely to produce new polyploid species on the archipelago, when the lineage already underwent polyploidization after diverging from its mainland sister lineage. Together, our results agree with previous conclusions that speciation on islands is similar to elsewhere (Takayama et al., 2018) and supports our hypothesis that polyploidy is a diversification trait on islands, with this trait being taxonomically more frequent in some lineages than others (Wood et al., 2009), e.g., Asteraceae (Figure 1). For example, Meudt et al. (2015) demonstrated that diversification in the hexaploid lineage of Veronica in New Zealand, the largest endemic lineage in New Zealand (Figure 1), is related to its decrease in genome FIGURE 3 | Estimated path analysis to explain the influence of polyploidy on size. This genome downsizing is also related to diversification the diversity of endemic island plant species. The model was fitted to 150 in mainland lineages of Veronica. A connection of genome lineages across four archipelagos for which we had published estimates of their stem age and chromosome numbers. Arrows are scaled proportional to downsizing in polyploid lineages has similarly been found in standardized effect sizes and only statistically significant pathways p( < 0.05) polyploid Cheirolophus (Asteraceae) on the Canary Islands are shown (e.g., direct effects of repeat colonization and source pool size on (Hidalgo et al., 2017). More generally, Kapralov and Filatov endemic diversity were tested but not significant). Gray text and arrows refer (2011) demonstrated a correlation of low genome size with to pathways that were statistically significant only on the Hawaiian (HI) or the species richness in endemic island genera. Thus, the relevance Juan Fernández (JF) Islands. of polyploidy for island diversification involves both lineage

Frontiers in Plant Science | www.frontiersin.org 8 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands features, such as the tendency to form polyploids, and aspects (Tables 1 and 2). Also, the age of New Zealand’s flora, as of island characteristics, such as habitat heterogeneity and estimated based on mean stem age per lineage (Table 2), does availability of pollinators. Islands also have many intrinsic not distinguish New Zealand from others due to the large features that may facilitate polyploid differentiation and variation in age. This large variation in ages of island lineages persistence, including small size, habitat heterogeneity and (Table 2) demonstrates that colonization has been successful proximity, and nutrient-rich soils associated with volcanic mostly independent of time of arrival. Thus, even old islands activity or maritime animals. are dispersal‐ or establishment-limited not niche-limited. Based It remains to be studied which aspects of polyploidy are on Carvajal-Endara et al. (2017) the establishment of a lineage critical for a given taxon and island and whether there is seems to be the bigger hurdle than dispersal. While our study a more general aspect that allows polyploids to disperse, focused on particular island systems that have been well studied, colonize, and/or diversify. Given advances in molecular the patterns of species diversification on islands that have biology, it seems feasible in the future to determine a emerged may be more general to other island systems (e.g., propensity to form polyploid species (Bomblies, 2020) and López-Alvarado et al., 2020). Other island systems would to detect ancient rounds of polyploidy (Tiley et al., 2018). be interesting to evaluate for these larger-scale patterns, as At the moment, polyploidy per se outside the island systems studies on their specific flora are becoming available (e.g., could not be considered in our analyses because of the Sardinia – López-Alvarado et al., 2020; Balearic Islands – difficulty to establish this in many cases (i.e., where the Rosselló and Castro, 2008). Indeed, Rice et al. (2019), in their chromosome number of a sister lineage was unknown) and global analysis of polyploid biogeography, find high levels of our observations that many of the lineages colonizing the polyploidy on several island systems, which mainly relate to island systems are already polyploid lineages of mainland the climatic conditions and predominance of perennial taxa. taxa [e.g., Coprosma (Rubiaceae)].Where there is high species richness within a genus, it seems the polyploid lineage is Polyploidy and Species Diversification on the one that colonizes the island systems, but overall these the Individual Islands examples remain few. One excellent example is the Hawaiian New Zealand is considered a continental island, because it silversword alliance, a recent radiation that includes ~30 separated from Gondwana 80 million years ago. Nevertheless, species in three endemic genera, which exhibit a variety of much of its flora, especially non-woody taxa, is considered to growth forms and exploit diverse habitats (Baldwin and have arrived by long distance-dispersal following large-scale Sanderson, 1998). The endemic Hawaiian species are all marine transgression during the Oligocene (Heenan and allotetraploid derivatives from diploid continental ancestors McGlone, 2019). Still, lineages on New Zealand included in (Carr, 1998; Barrier et al., 1999). Our result of increased our analyses are on average older than those of the other diversity on the island systems based on a higher ploidy island systems (Table 2). Sanmartín et al. (2007) inferred most level of the island lineage compared to the mainland sister of the colonization events to have occurred from Australia, lineage, nevertheless, allows us to focus on the question of and our survey of available phylogenies shows that many whether polyploidy is a dispersal or colonization trait. New Zealand genera are part of larger Southern Hemisphere Crawford et al. (2009) already queried whether the colonizers lineages. New Zealand has the largest flora of those studied themselves were polyploid or if polyploidy evolved in situ, here with a high number of endemics and several large radiations which motivated our first prediction (P1) in our conceptual involving polyploid formation (e.g., Veronica; Meudt et al., model. Our literature survey provides many examples of 2015, Table 1). The flora is well-studied phylogenetically and such cases to be studied in the future under different chromosomally with genome size measurements increasing evolutionary scenarios, with and without further diversification, recently, which makes New Zealand an excellent example for with or without further changes in ploidy level, and in studies on the importance of polyploidy. different geological and temporal settings. The Canary Islands are characterized among the five archipelagos as the one closest to a continent, which may Polyploidy and Diversification on Islands – explain its low endemicity despite having the highest diversity Generalities per square kilometer of all island systems except Juan Fernández Islands have complex geological histories, and generally (Table 1). Indeed, nearby mainland Europe or Africa seems New Zealand is considered to be a “continental island” as to be the ancestral areas for most Canary Island genera and compared to the other “oceanic islands” that we studied here species (Sanmartín et al., 2008). This diversity and the ease (Weigelt et al., 2013). However, our analyses showed that this to access the island system explains why the Canary Islands was not a meaningful distinction for determining factors have become such an important natural laboratory for influencing diversification as results from all island systems evolutionary botanists with studies investigating patterns of were generally similar (few island effects, and none that dispersal and evolution on the archipelago, such as the evolution distinguished New Zealand from the other four archipelagos). of woodiness (Böhle et al., 1996; Schüßler et al., 2019), Thus, despite the fact that New Zealand has not been included breeding system (Soto-Trejo et al., 2013), and photosynthetic with other studies of island polyploid speciation, we find pathways (Mort et al., 2007). Most important in our context similarities among all the island systems included here in terms is the intensive karyological study of the archipelago (e.g., of the contribution of polyploidy to species diversification Suda et al., 2005; Table 1). Interestingly, a number of studies

Frontiers in Plant Science | www.frontiersin.org 9 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

(e.g., Allan et al., 2004) demonstrate that Canary Island taxa island system. The high number of polyploids in the ancestral occupy similar habitats as compared to their continental lineages of some Juan Fernández endemics, such as Dendroseris relatives, suggesting that their ancestors might have been and Erigeron (Asteraceae), therefore suggests that polyploidy pre-adapted to occupying particular habitats. is a dispersal or colonization trait rather than a diversification Polyploidy seems to have a larger impact on diversification trait for the flora of Juan Fernández. Two factors could on New Zealand and the Canary Islands compared to the explain this lack of polyploid speciation on Juan Fernández other island systems, for example in the largest radiation on Islands: their small size providing little space for the avoidance the Canary Islands, the Aeonium alliance (Figure 1), or Sideritis of inter-cytotype gene flow or the young age of the lineages (Lamiaceae; Raskina et al., 2008). The suggestion that polyploidy compared to that of the other islands (Table 2). Given more seems to be less common in Asteraceae on the Canary Islands time the intraspecific variation in ploidy of several species compared to other island systems (Crawford et al., 2009) only on the islands [Stuessy et al., 2017; e.g., in Eryngium holds for the two largest genera of Asteraceae on the island bupleuroides (Asteraceae); Figure 1] could translate into new system (Argyranthemum, Sonchus alliance), not generally across polyploid species. the family. More detailed analysis on the distribution of polyploids The Galápagos have not been included in our analyses, on New Zealand and the Canary Islands will be necessary to despite the importance of this island system for the history determine whether, for example, the larger altitudinal range of evolution and biogeography, because its flora has been poorly on these two archipelagos promote diversification by polyploidy. studied. Its flora, like its fauna, is considered to be derived This should be analyzed in connection with the report of the from South America (Darwin, 1839; Elisens, 1992), although low genome size of the Macaronesian flora in relation with exceptions occur (e.g., Andrus et al., 2009). The Galápagos the rest of the world (Suda et al., 2005). flora is only a third the size of the Canary Islands flora despite The Hawaiian Islands are the island system farthest away similar area, possibly due to its dryness, which could also from a continent, and thus its genera are mostly considered explain the low number of endemic fern species (Table 1). to have colonized the island system only once (Funk and Polyploid speciation, however, has been shown to occur in Wagner, 1995; Keeley and Funk, 2011). Following from this the genus Pectis (Asteraceae, Hansen et al., 2016). In Scalesia it also has the highest number of endemics (Table 1). Ancestors (Asteraceae), the largest radiation on the archipelago, the species of Hawaiian taxa are derived from diverse geographical regions are considered to be tetraploid but have not further diversified including North America, Australia, Asia, Africa, and other in ploidy level (Eliasson, 1974; Fernández-Mazuecos et al., 2020). island systems in the Pacific (e.g., New Zealand;Keeley and The discussion of these five island systems already demonstrates Funk, 2011; Price and Wagner, 2018). The considerable age that there are as many idiosyncratic patterns as there are island of some of its lineages is likely caused by the influence of systems, but polyploidy seems to play a role in many of them. early colonization of now succumbed islands of the archipelago For example, Rosselló and Castro (2008) found a large frequency (García-Verdugo et al., 2019), while most others represent more of neopolyploidization events among the endemic plants of the recent arrivals (Price and Clague, 2002). Despite great interest Balearic Islands, whereas Sun and Stuessy (1998) recorded in the flora of the island system, it is only poorly investigated chromosomal stasis in the flora of Ullung Islands. Studies of karyologically (Table 1). Thus, the hypothesis of chromosomal additional island systems that vary in their locations and floristic stasis of lineages on the Hawaiian Islands (Stuessy and Crawford, complexities would be beneficial for enhancing our understanding 1998) is founded on few, though important, cases. Nevertheless, of endemic polyploid diversification. the data available suggest a lower importance of polyploidy than in New Zealand and the Canary Islands, chromosomal stasis thus being peculiar to the Hawaiian Islands. One notable OUTLOOK AND FUTURE DIRECTIONS pattern for the Hawaiian Islands is the formation of dysploid series (Carr, 1998), which has also been noted on rare occasions We see the study of polyploidy on islands as an exciting in the Canary Islands (Sideritis, Barber et al., 2000) and avenue to further our understanding of polyploid species New Zealand (Veronica, Wagstaff and Garnock-Jones, 1998). diversification. Our comparative study was limited to data The Juan Fernández Islands are the smallest archipelago that were available for the island systems under study and but the flora has been intensively investigated floristically we were surprised that some of these “classic” island systems (Greimler et al., 2013) and evolutionarily (Takayama et al., remain poorly known chromosomally and phylogenetically, 2018), including a number of karyological studies but so especially the Galápagos. While genomic studies have far no measurements of genome sizes (Table 1). The closest revolutionized our ideas of polyploid genome dynamics (e.g., continental source is South America, which seems to be the Bomblies, 2020), there is much value in continuing to generate ancestral area for several Juan Fernández genera and species chromosome numbers and dated phylogenies for native island (Stuessy et al., 2017). Despite the fact that some genera endemics. Because island radiations are often young and/or have species with different ploidy levels on different islands, rapid on an evolutionary timescale, in many cases acquiring there is no clear case of polyploid origin there because those a well-resolved phylogeny for a group of species based on cases are either based on separate colonization events or single or few gene sequences has been challenging (e.g., Meudt have not been studied phylogenetically. Also, Stuessy et al. (2017) and Simpson, 2006; Knope et al., 2012; Vitales et al., 2014). state only two possible cases of polyploid origins on this Newer methodologies that take advantage of next-generation

Frontiers in Plant Science | www.frontiersin.org 10 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands sequencing methods should be helpful in this regard (Larridon FUNDING et al., 2020). Similarly, overseas sister lineages need to be investigated to understand the context of island We thank the Marsden Fund, Royal Society Te Apārangi diversification. One additional limitation to our study was (LCR1702) for research funding focused on polyploidy in the the lack of data for genome size estimates, either to help native New Zealand flora to WL, HM, and JT; the Gatsby resolve ploidy levels or to analyze its effect on polyploid Charitable Foundation (Grant Number GAT2962) to AT for species diversification on the island systems. Given that genome funding focused on evolutionary diversification, and the German downsizing is considered to be an important part of polyploid Science Foundation (DFG; project AL632/16-1) for travel support evolution (Soltis et al., 2016), we see this factor as a potentially to New Zealand for DA. important player in facilitating establishment on islands, as mentioned previously for Veronica (Meudt et al., 2015). Additionally, future studies should further compare the effect ACKNOWLEDGMENTS of polyploidy on colonization and diversification among woody vs. herbaceous, dry vs. fleshy fruited, and selfing vs. outcrossing Thanks to Clyde Imada and Cliff Morden for making data lineages (Vamosi et al., 2007) for a more complete picture from Hawaiian species available, and to Patrick Brownsey for of island diversification. providing information regarding some fern taxa.

DATA AVAILABILITY STATEMENT SUPPLEMENTARY MATERIAL

The original contributions presented in the study are included The Supplementary Material for this article can be found online in the article/Supplementary Material, further inquiries can at: https://www.frontiersin.org/articles/10.3389/fpls.2021.637214/ be directed to the corresponding authors. full#supplementary-material

Supplementary File 1 | The data file analyzed in this study, which includes the following information from 150 island lineages from New Zealand (n = 98), AUTHOR CONTRIBUTIONS Canary Islands (n = 23), Hawaiian Islands (n = 23), and Juan Fernández Islands (n = 6): lineage (genus) name, number of endemic species, stem age, number HMM, JAT, AJT, DCA, and WGL designed the study and of ploidy levels, whether ploidy level differs from closest non-island sister group, compiled the data, with assistance from SCN, AJB and JI. monophyly on the island, and number of species in closest non-island sister group. AJT and JI completed statistical analyses. All authors contributed to drafting the manuscript and developing the final version Supplementary File 2 | References for the 150 lineages included in the and are accountable for the contents of the work. statistical analyses (with dated phylogenies) grouped by island system.

REFERENCES Böhle, U. R., Hilger, H. H., and Martin, W. F. (1996). Island colonization and evolution of the insular woody habit in Echium L. (Boraginaceae). Proc. Allan, G. J., Francisco-Ortega, J., Santos-Guerra, A., Boerner, E., and Zimmer, E. A. Natl. Acad. Sci. 93, 11740–11745. doi: 10.1073/pnas.93.21.11740 (2004). Molecular phylogenetic evidence for the geographic origin and Bomblies, K. (2020). When everything changes at once: finding a new normal classification of Canary IslandLotus (Fabaceae: Loteae). Mol. Phylogenet. after genome duplication. Proc. R. Soc. B Biol. Sci. 287:20202154. doi: 10.1098/ Evol. 32, 123–138. doi: 10.1016/j.ympev.2003.11.018 rspb.2020.2154 Andrus, N., Tye, A., Nesom, G., Bogler, D., Lewis, C., Noyes, R., et al. (2009). Brandt, A., Tanentzap, A., Leopold, D., Heenan, P., Fukami, T., and Lee, W. Phylogenetics of Darwiniothamnus (Asteraceae: Astereae) – molecular evidence (2016). Precipitation alters the strength of evolutionary priority effects in for multiple origins in the endemic flora of the Galápagos Islands. J. Biogeogr. forest community assembly of pteridophytes and angiosperms. J. Ecol. 104, 36, 1055–1069. doi: 10.1111/j.1365-2699.2008.02064.x 1673–1681. doi: 10.1111/1365-2745.12640 Arechavaleta, M., Rodríguez, S., Zurita, N., and García, A. (2010). Lista de especies Breitwieser, I., Brownsey, P. J., Garnock-Jones, P. J., Perrie, L. R., and Wilton, A. D. silvestres de Canarias. Hongos, plantas y animales terrestres. 2009. Tenerife: (2012). Phylum Tracheophyta: Vascular plants. Christchurch: Canterbury Gobierno de Canarias. University Press. Baker, H. G. (1967). Support for Baker’s law ‐ as a rule. Evolution 21, 853–856. Carlquist, S. (1974). Island biology. New York: Columbia University Press. doi: 10.1111/j.1558-5646.1967.tb03440.x Carr, G. D. (1998). “Chromosome evolution and speciation in Hawaiian flowering Baldwin, B. G., and Sanderson, M. J. (1998). Age and rate of diversification plants” in Evolution and speciation of island plants. eds. T. F. Stuessy and of the Hawaiian silversword alliance (Compositae). Proc. Natl. Acad. Sci. M. Ono (Cambridge: Cambridge University Press), 5–47. 95, 9402–9406. doi: 10.1073/pnas.95.16.9402 Carracedo, J. C., and Troll, V. R. (2016). The geology of the Canary Islands. Barber, J. C., Ortega, J. F., Santos-Guerra, A., Marrero, A., and Jansen, R. K. London: Elsevier. (2000). Evolution of endemic Sideritis (Lamiaceae) in Macaronesia: insights Carvajal-Endara, S., Hendry, A. P., Emery, N. C., and Davies, T. J. (2017). Habitat from a chloroplast DNA restriction site analysis. Syst. Bot. 25, 633–647. filtering not dispersal limitation shapes oceanic island floras: species assembly doi: 10.2307/2666725 of the Galápagos archipelago. Ecol. Lett. 20, 495–504. doi: 10.1111/ele.12753 Barrier, M., Baldwin, B. G., Robichaux, R. H., and Purugganan, M. D. (1999). Caujapé-Castells, J., García-Verdugo, C., Marrero-Rodríguez, Á., Interspecific hybrid ancestry of a plant adaptive radiation: allopolyploidy Fernández-Palacios, J. M., Crawford, D. J., and Mort, M. E. (2017). Island of the Hawaiian silversword alliance (Asteraceae) inferred from floral homeotic ontogenies, syngameons, and the origins and evolution of genetic diversity gene duplications. Mol. Biol. Evol. 16, 1105–1113. doi: 10.1093/oxfordjournals. in the Canarian endemic flora. Perspect. Plant Ecol. Evol. Syst. 27, 9–22. molbev.a026200 doi: 10.1016/j.ppees.2017.03.003

Frontiers in Plant Science | www.frontiersin.org 11 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

Crawford, D. J., and Archibald, J. K. (2017). Island floras as model systems Hidalgo, O., Vitales, D., Vallès, J., Garnatje, T., Siljak-Yakovlev, S., Leitch, I. J., for studies of plant speciation: prospects and challenges. J. Syst. Evol. 55, et al. (2017). Cytogenetic insights into an oceanic island radiation: the 1–15. doi: 10.1111/jse.12234 dramatic evolution of pre-existing traits in Cheirolophus (Asteraceae: Cardueae: Crawford, D. J., Lowrey, T. K., Anderson, G. J., Bernardello, G., Santos-Guerra, A., Centaureinae). Taxon 66, 146–157. doi: 10.12705/661.8 and Stuessy, T. F. (2009). “Genetic diversity in Asteraceae endemic to ocean Imada, C. T. (2012). Hawaiian Native and Naturalized Vascular Plants Checklist islands: Baker’s law and polyploidy” in Systematics, evolution, and biogeography (December 2012 update). Bishop Musem Technical Report 60. of compositae. eds. V. A. Funk, T. F. Stuessy, A. Susanna and R. J. Bayer Jaramillo Díaz, P., Guézou, A., Mauchamp, A., and Tye, A. (2017). “CDF (Vienna: International Association of Plant Taxonomists), 101–113. checklist of Galapagos ferns and related groups ‐ FCD Lista de especies Crawford, D. J., and Stuessy, T. F. (1997). “Plant speciation on oceanic islands” de Helechos y grupos relacionados Galápagos” in Charles Darwin Foundation in Evolution and diversification of land plants. eds. K. Iwatsuki and P. H. Raven Galapagos species checklist ‐ Lista de Especies de Galápagos de la Foundation (Tokyo: Springer-Verlag), 249–267. Charles Darwin. eds. F. Bungartz, H. Herrera, P. Jaramillo, N. Tirado, Darwin, C. R. (1839). Narrative of the surveying voyages of His Majesty’s Ships G. Jiménez-Uzcátegui, D. Ruiz, et al (Puerto Ayora, Galapagos: Charles Adventure and Beagle between the years 1826 and 1836, describing their Darwin Foundation / Fundación Charles Darwin). examination of the southern shores of South America, and the Beagle’s Jaramillo Díaz, P., Guézou, A., Mauchamp, A., and Tye, A. (2018). “CDF checklist circumnavigation of the globe. Journal and remarks. 1832-1836. Vol. 3. London: of Galapagos flowering plants ‐ FCD Lista de especies de Plantas con flores Henry Colburn. Galápagos” in Charles Darwin Foundation Galapagos species checklist ‐ Dawson, M. I. (2000). Index of chromosome numbers of indigenous New Zealand Lista de Especies de Galápagos de la Foundation Charles Darwin. eds. spermatophytes. N. Z. J. Bot. 38, 47–150. doi: 10.1080/0028825X.2000.9512673 F. Bungartz, H. Herrera, P. Jaramillo, N. Tirado, G. Jiménez-Uzcátegui, Dawson, M. I. (2008). Index of chromosome numbers of indigenous New Zealand D. Ruiz, et al (Puerto Ayora, Galapagos: Charles Darwin Foundation / Fundación vascular plants [Online]. Landcare Research, New Zealand. Available at: Charles Darwin). http://www.landcareresearch.co.nz (Accessed February 23, 2021). Jin, Y., and Qian, H. (2019). V.PhyloMaker: an R package that can generate Eliasson, U. (1974). Studies in Galápagos plants XIV. The genus Scalesia Arn. very large phylogenies for vascular plants. Ecography 42, 1353–1359. doi: Opera Bot. 36, 1–117. 10.1111/ecog.04434 Elisens, W. J. (1992). Genetic divergence in Galvezia (Scrophulariaceae): Kapralov, M. V., and Filatov, D. A. (2011). Does large genome size limit speciation evolutionary and biogeographic relationships among south American and in endemic island floras?J. Bot. 2011:458684. doi: 10.1155/2011/458684 Galápagos species. Am. J. Bot. 79, 198–206. doi: 10.1002/j.1537-2197.1992. Keeley, S. C., and Funk, V. A. (2011). “Origin and evolution of Hawaiian tb13638.x endemics: new patterns revealed by molecular phylogenetic studies” in The Emerson, B. (2002). Evolution on oceanic islands: molecular phylogenetic biology of island floras. eds. D. Bramwell and J. Caujapé-Castells (Cambridge: approaches to understanding pattern and process. Mol. Ecol. 11, 951–966. Cambridge University Press), 57–88. doi: 10.1046/j.1365-294x.2002.01507.x Kellogg, E. A. (2016). Has the connection between polyploidy and diversification Fernández-Mazuecos, M., Vargas, P., McCauley, R. A., Monjas, D., Otero, A., actually been tested? Curr. Opin. Plant Biol. 30, 25–32. doi: 10.1016/j. Chaves, J. A., et al. (2020). The radiation of Darwin’s giant daisies in pbi.2016.01.002 the Galápagos Islands. Curr. Biol. 30, 4989–4998.e7. doi: 10.1016/j. Kim, S. -C., McGowen, M. R., Lubinsky, P., Barber, J. C., Mort, M. E., and cub.2020.09.019 Santos-Guerra, A. (2008). Timing and tempo of early and successive adaptive Frankham, R. (1997). Do island populations have less genetic variation than radiations in Macaronesia. PLoS One 3:e2139. doi: 10.1371/journal.pone.0002139 mainland populations? Heredity 78, 311–327. doi: 10.1038/hdy.1997.46 Knope, M. L., Morden, C. W., Funk, V. A., and Fukami, T. (2012). Area and Funk, V. A., and Wagner, W. L. (1995). Hawaiian biogeography: Evolution on the rapid radiation of Hawaiian Bidens (Asteraceae). J. Biogeogr. 39, 1206–1216. a hot spot archipelago. Washington, DC: Smithsonian Institution Press. doi: 10.1111/j.1365-2699.2012.02687.x Garamszegi, L. Z., and Møller, A. P. (2010). Effects of sample size and intraspecific Kuo, L. -Y., Ebihara, A., Shinohara, W., Rouhan, G., Wood, K. R., Wang, C. -N., variation in phylogenetic comparative studies: a meta-analytic review. Biol. et al. (2016). Historical biogeography of the fern genus Deparia (Athyriaceae) Rev. 85, 797–805. doi: 10.1111/j.1469-185X.2010.00126.x and its relation with polyploidy. Mol. Phylogenet. Evol. 104, 123–134. doi: García-Verdugo, C., Caujapé-Castells, J., and Sanmartín, I. (2019). Colonization 10.1016/j.ympev.2016.08.004 time on island settings: lessons from the Hawaiian and Canary Island floras. Larridon, I., Villaverde, T., Zuntini, A. R., Pokorny, L., Brewer, G. E., Epitawalage, N., Bot. J. Linn. Soc. 191, 155–163. doi: 10.1093/botlinnean/boz044 et al. (2020). Tackling rapid radiations with targeted sequencing. Front. García-Verdugo, C., Sajeva, M., La Mantia, T., Harrouni, C., Msanda, F., and Plant Sci. 10:1655. doi: 10.3389/fpls.2019.01655 Caujapé-Castells, J. (2015). Do island plant populations really have lower Lee, W., Tanentzap, A., and Heenan, P. (2012). Plant radiation history affects genetic variation than mainland populations? Effects of selection and community assembly: evidence from the New Zealand alpine. Biol. Lett. 8, distribution range on genetic diversity estimates. Mol. Ecol. 24, 726–741. 558–561. doi: 10.1098/rsbl.2011.1210 doi: 10.1111/mec.13060 Lefcheck, J. S. (2016). piecewiseSEM: piecewise structural equation modeling Garnock-Jones, P. J. (2014). Evidence-based review of the taxonomic status of in R for ecology, evolution, and systematics. Methods Ecol. Evol. 7, 573–579. New Zealand’s endemic seed plant genera. N. Z. J. Bot. 52, 163–212. doi: doi: 10.1111/2041-210X.12512 10.1080/0028825X.2014.902854 Linder, H. P., and Barker, N. P. (2014). Does polyploidy facilitate long-distance Greimler, J., López-Sepulveda, P., Reiter, K., Baeza, C., Peñailillo, P., Ruiz, E., dispersal? Ann. Bot. 113, 1175–1183. doi: 10.1093/aob/mcu047 et al. (2013). Vegetation of Alejandro Selkirk island (Isla Masafuera), Juan López-Alvarado, J., Mameli, G., Farris, E., Susanna, A., Filigheddu, R., and Fernández archipelago. Chile. Pac. Sci. 67, 267–282. doi: 10.2984/67.2.9 Garcia-Jacas, N. (2020). Islands as a crossroad of evolutionary lineages: a Hansen, D. R., Jansen, R. K., Sage, R. F., Villaseñor, J. L., and Simpson, B. B. case study of Centaurea sect. Centaurea (Compositae) from Sardinia

(2016). Molecular phylogeny of Pectis (Tageteae, Asteraceae), a C4 genus of (Mediterranean Basin). PLoS One 15:e0228776. doi: 10.1371/journal. the Neotropics, and its sister genus Porophyllum. Lundellia 19, 6–38. doi: pone.0228776 10.25224/1097-993X-19.1.6 MacArthur, R. H., and Wilson, E. O. (1967). The theory of island biogeography. Harbaugh, D. (2008). Polyploid and hybrid origins of Pacific island sandalwoods Princeton, New Jersey: Princeton University Press. (Santalum, Santalaceae) inferred from low-copy nuclear and flow cytometry Mayr, E. (1967). The challenge of island faunas.Australian Nat. Hist. 15, data. Int. J. Plant Sci. 169, 677–685. doi: 10.1086/533610 369–374. Heenan, P. B., and McGlone, M. S. (2019). Cenozoic formation and colonisation McGlone, M. S., Duncan, R. P., and Heenan, P. B. (2001). Endemism, species history of the New Zealand vascular flora based on molecular clock dating selection and the origin and distribution of the vascular plant flora of of the plastid rbcL gene. N. Z. J. Bot. 57, 204–226. doi: 10.1080/0028825X. New Zealand. J. Biogeogr. 28, 199–216. doi: 10.1046/j.1365-2699.2001.00525.x 2019.1632356 Meudt, H. M., Rojas‐Andrés, M., Prebble, J. M., Low, E., Garnock-Jones, P. J., Herben, T., Suda, J., and Klimešová, J. (2017). Polyploid species rely on vegetative and Albach, D. C. (2015). Is genome downsizing associated with reproduction more than diploids: a re-examination of the old hypothesis. diversification in polyploid lineages of Veronica? Bot. J. Linn. Soc. 178, Ann. Bot. 120, 341–349. doi: 10.1093/aob/mcx009 243–266. doi: 10.1111/boj.12276

Frontiers in Plant Science | www.frontiersin.org 12 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

Meudt, H. M., and Simpson, B. B. (2006). The biogeography of the austral, Schönberger, I., Wilton, A. D., Boardman, K. F., Breitwieser, I., de Lange, P. J., subalpine genus Ourisia (Plantaginaceae) based on molecular phylogenetic de Pauw, B., et al. (2019). Checklist of the New Zealand Flora – Seed plants. evidence: south American origin and dispersal to New Zealand and Tasmania. Manaaki Whenua-Landcare Research: Lincoln. Biol. J. Linn. Soc. 87, 479–513. doi: 10.1111/j.1095-8312.2006.00584.x Schüßler, C., Bräuchler, C., Reyes-Betancort, J. A., Koch, M. A., and Thiv, M. Mort, M. E., Solits, D. E., Soltis, P. S., Santos-Guerra, A., and Francisco-Ortega, J. (2019). Island biogeography of the Macaronesian Gesnouinia and Mediterranean (2007). Physiological evolution and association between physiology and growth Soleirolia (Parietarieae, Urticaceae) with implications for the evolution of form in Aeonium (Crassulaceae). Taxon 56, 453–464. doi: 10.1002/tax.562016 insular woodiness. Taxon 68, 537–556. doi: 10.1002/tax.12061 Moura, R. F., Queiroga, D., Vilela, E., and Moraes, A. P. (2020). Polyploidy Smith, S. A., and Brown, J. W. (2018). Constructing a broadly inclusive seed and high environmental tolerance increase the invasive success of plants. plant phylogeny. Am. J. Bot. 105, 302–314. doi: 10.1002/ajb2.1019 J. Plant Res. 134, 105–114. doi: 10.1007/s10265-020-01236-6 Soltis, D. E., Albert, V. A., Leebens-Mack, J., Bell, C. D., Paterson, A. H., Murray, B. G., and de Lange, P. J. (2011). “Chromosomes and evolution in Zheng, C. F., et al. (2009). Polyploidy and angiosperm diversification.Am. New Zealand endemic angiosperms and gymnosperms” in The biology of J. Bot. 96, 336–348. doi: 10.3732/ajb.0800079 island floras. eds. D. Bramwell and J. Caujapé-Castells (Cambridge: Cambridge Soltis, P. S., Liu, X., Marchant, D. B., Visger, C. J., and Soltis, D. E. (2014). University Press), 265–283. Polyploidy and novelty: Gottlieb’s legacy. Philos. Trans. R. Soc. Lond. B Pellicer, J., and Leitch, I. J. (2020). The plant DNA C-values database (release 369:20130351. doi: 10.1098/rstb.2013.0351 7.1): an updated online repository of plant genome size data for comparative Soltis, D. E., Visger, C. J., Marchant, D. B., and Soltis, P. S. (2016). Polyploidy: pitfalls studies. New Phytol. 226, 301–305. doi: 10.1111/nph.16261 and paths to a paradigm. Am. J. Bot. 103, 1146–1166. doi: 10.3732/ajb.1500501 Pinheiro, J., Bates, D., DebRoy, S., and Sarkar, D. (2014). nlme: linear and Soto-Trejo, F., Kelly, J. K., Archibald, J. K., Mort, M. E., Santos-Guerra, A., nonlinear mixed effects models. R package version 3.1-117.R Core Team and Crawford, D. J. (2013). The genetics of self-compatibility and associated (2014). Available at: https://CRAN.R-project.org/package=nlme (Accessed floral characters in Tolpis (Asteraceae) in the Canary Islands. Int. J. Plant February 23, 2021). Sci. 174, 171–178. doi: 10.1086/668788 Price, J. P. (2004). Floristic biogeography of the Hawaiian islands: influences Stebbins, G. L. (1947). Types of polyploids: their classification and significance. of area, environment and paleogeography. J. Biogeogr. 31, 487–500. doi: Adv. Genet. 1, 403–429. doi: 10.1016/s0065-2660(08)60490-3 10.1046/j.0305-0270.2003.00990.x Stuessy, T. F., and Crawford, D. J. (1998). “Chromosomal stasis during speciation in Price, J. P., and Clague, D. A. (2002). How old is the Hawaiian biota? Geology angiosperms of oceanic islands” in Evolution and speciation of island plants. eds. and phylogeny suggest recent divergence. Proc. R. Soc. Lond. B Biol. Sci. T. F. Stuessy and M. Ono (Cambridge: Cambridge University Press), 307–324. 269, 2429–2435. doi: 10.1098/rspb.2002.2175 Stuessy, T. F., Crawford, D. J., López-Sepúlveda, P., Baeza, C. M., and Ruiz, E. A. Price, J. P., and Wagner, W. L. (2018). Origins of the Hawaiian flora: phylogenies (eds.) (2017). Plants of oceanic islands: Evolution, biogeography, and conservation and biogeography reveal patterns of long-distance dispersal. J. Syst. Evol. of the flora of the Juan Fernández (Robinson Crusoe) archipelago. (Cambridge: 56, 600–620. doi: 10.1111/jse.12465 Cambridge University Press). Raskina, O., Barber, J. C., Nevo, E., and Belyayev, A. (2008). Repetitive DNA Stuessy, T. F., Crawford, D. J., Marticorena, C., and Rodríguez, R. (1998). and chromosomal rearrangements: speciation-related events in plant genomes. “Island biogeography of angiosperms of the Juan Fernandez archipelago” Cytogenet. Genome Res. 120, 351–357. doi: 10.1159/000121084 in Evolution and speciation of island plants. eds. T. F. Stuessy and M. Ono Rice, A., Šmarda, P., Novosolov, M., Drori, M., Glick, L., Sabath, N., et al. (Cambridge: Cambridge University Press), 121–140. (2019). The global biogeography of polyploid plants.Nat. Ecol. Evol. 3, Stuessy, T. F., Takayama, K., López-Sepúlveda, P., and Crawford, D. J. (2014). 265–273. doi: 10.1038/s41559-018-0787-9 Interpretation of patterns of genetic variation in endemic plant species of Rivas-Torres, G. F., Benítez, F. L., Rueda, D., Sevilla, C., and Mena, C. F. oceanic islands. Bot. J. Linn. Soc. 174, 276–288. doi: 10.1111/boj.12088 (2018). A methodology for mapping native and invasive vegetation coverage Suda, J., Kyncl, T., and Jarolímová, V. (2005). Genome size variation in in archipelagos: an example from the Galápagos Islands. Prog. Phys. Geogr. Macaronesian angiosperms: forty percent of the Canarian endemic flora 42, 83–111. doi: 10.1177/0309133317752278 completed. Plant Syst. Evol. 252, 215–238. doi: 10.1007/s00606-004-0280-6 Robertson, K., Goldberg, E. E., and Igić, B. (2011). Comparative evidence for Sun, B. -Y., and Stuessy, T. F. (1998). “Preliminary observations on the evolution the correlated evolution of polyploidy and self-compatibility in Solanaceae. of endemic angiosperms” in Evolution and speciation of island plants. eds. Evolution 65, 139–155. doi: 10.1111/j.1558-5646.2010.01099.x T. F. Stuessy and M. Ono (Cambridge: Cambridge University Press), 181–202. Rodriguez, D. J. (1996). A model for the establishment of polyploidy in plants. Symonds, M. R. E., and Blomberg, S. P. (2014). “A primer on phylogenetic Am. Nat. 147, 33–46. generalised least squares” in Modern phylogenetic comparative methods and Rosche, C., Hensen, I., Mráz, P., Durka, W., Hartmann, M., and Lachmuth, S. their application in evolutionary biology. ed. L. Z. Garamszegi (Berlin, (2017). Invasion success in polyploids: the role of inbreeding in the contrasting Heidelberg: Springer), 105–130. colonization abilities of diploid versus tetraploid populations of Centaurea Takayama, K., Crawford, D. J., López-Sepúlveda, P., Greimler, J., and Stuessy, T. F. stoebe s.l. J. Ecol. 105, 425–435. doi: 10.1111/1365-2745.12670 (2018). Factors driving adaptive radiation in plants of oceanic islands: a Rosenzweig, M. L. (1995). Species diversity in space and time. New York: case study from the Juan Fernández archipelago. J. Plant Res. 131, 469–485. Cambridge University Press. doi: 10.1007/s10265-018-1023-z Rosselló, J. A., and Castro, M. (2008). Karyological evolution of the angiosperm Tanentzap, A. J., Brandt, A. J., Smissen, R. D., Heenan, P. B., Fukami, T., and endemic flora of the Balearic Islands. Taxon 57, 259–273. doi: 10.2307/ Lee, W. G. (2015). When do plant radiations influence community assembly? 25065967 The importance of historical contingency in the race for niche space. New Sanmartín, I., Van Der Mark, P., and Ronquist, F. (2008). Inferring dispersal: Phytol. 207, 468–479. doi: 10.1111/nph.13362 a Bayesian approach to phylogeny-based island biogeography, with special Te Beest, M., Le Roux, J. J., Richardson, D. M., Brysting, A. K., Suda, J., reference to the Canary Islands. J. Biogeogr. 35, 428–449. doi: 10.1111/j.1365- Kubesova, M., et al. (2012). The more the better? The role of polyploidy 2699.2008.01885.x in facilitating plant invasions. Ann. Bot. 109, 19–45. doi: 10.1093/aob/mcr277 Sanmartín, I., Wanntorp, L., and Winkworth, R. C. (2007). West wind drift revisited: Tiley, G. P., Barker, M. S., and Burleigh, J. G. (2018). Assessing the performance testing for directional dispersal in the southern hemisphere using event-based of Ks plots for detecting ancient whole genome duplications. Genome Biol. tree fitting.J. Biogeogr. 34, 398–416. doi: 10.1111/j.1365-2699.2006.01655.x Evol. 10, 2882–2898. doi: 10.1093/gbe/evy200 Savolainen, V., Anstett, M. -C., Lexer, C., Hutton, I., Clarkson, J. J., Norup, M. V., The Plant List (2013). Published on the Internet. Available at:http://www. et al. (2006). Sympatric speciation in palms on an oceanic island. Nature theplantlist.org/ (Accessed February 23, 2021). 441, 210–213. doi: 10.1038/nature04566 Vamosi, J. C., Goring, S. J., Kennedy, B. F., Mayberry, R. J., Moray, C. M., Scarrow, M., Wang, Y., and Sun, G. (2020). Molecular regulatory mechanisms Neame, L. A., et al. (2007). Pollination, floral display, and the ecological underlying the adaptability of polyploid plants. Biol. Rev. doi:10.1111/brv.12661 correlates of polyploidy. Funct. Ecosyst. Communities 1, 1–9. Scholl, J. P., and Wiens, J. J. (2016). Diversification rates and species richness Vamosi, J. C., Magallón, S., Mayrose, I., Otto, S. P., and Sauquet, H. (2018). across the tree of life. Proc. R. Soc. B Biol. Sci. 283:20161334. doi: 10.1098/ Macroevolutionary patterns of flowering plant speciation and extinction. rspb.2016.1334 Annu. Rev. Plant Biol. 69, 685–706. doi: 10.1146/annurev-arplant-042817-040348

Frontiers in Plant Science | www.frontiersin.org 13 March 2021 | Volume 12 | Article 637214 Meudt et al. Polyploidy on Islands

Vargas, P., Arjona, Y., Nogales, M., and Heleno, R. H. (2015). Long-distance Wood, T., (2009). The frequency of polyploid speciation in vascular plants. dispersal to oceanic islands: success of plants with multiple diaspore Proc. Natl. Acad. Sci. 106, 13875–13879. doi: 10.1073/pnas.0811 specializations. AoB PLANTS 7:plv073. doi: 10.1093/aobpla/plv073 575106 Vitales, D., Garnatje, T., Pellicer, J., Vallès, J., Santos-Guerra, A., and Sanmartín, I. (2014). The explosive radiation ofCheirolophus (Asteraceae, Cardueae) in Conflict of Interest: The authors declare that the research was conducted in Macaronesia. BMC Evol. Biol. 14:118. doi: 10.1186/1471-2148-14-118 the absence of any commercial or financial relationships that could be construed Wagner, W. L., Herbst, D. R., and Lorence, D. H. (2005). Flora of the Hawaiian as a potential conflict of interest. Islands website [Online]. Available at: http://botany.si.edu/pacificislandbiodiversity/ hawaiianflora/index.htm (Accessed January 1, 2020). Copyright © 2021 Meudt, Albach, Tanentzap, Igea, Newmarch, Brandt, Lee and Wagstaff, S. J., and Garnock-Jones, P. J. (1998). Evolution and biogeography Tate. This is an open-access article distributed under the terms of theCreative of the Hebe complex (Scrophulariaceae) inferred from ITS sequences. N. Commons Attribution License (CC BY). The use, distribution or reproduction in Z. J. Bot. 36, 425–437. doi: 10.1080/0028825X.1998.9512581 other forums is permitted, provided the original author(s) and the copyright owner(s) Weigelt, P., Jetz, W., and Kreft, H. (2013). Bioclimatic and physical characterization are credited and that the original publication in this journal is cited, in accordance of the world’s islands. Proc. Natl. Acad. Sci. 110, 15307–15312. doi: 10.1073/ with accepted academic practice. No use, distribution or reproduction is permitted pnas.1306309110 which does not comply with these terms.

Frontiers in Plant Science | www.frontiersin.org 14 March 2021 | Volume 12 | Article 637214 MASSEY UNIVERSITY MASSEY RESEARCH ONLINE http://mro.massey.ac.nz/

Massey Documents by Type Journal Articles

Polyploidy on Islands: Its Emergence and Importance for Diversification.

Meudt, HM

2021

10/09/2021 - Downloaded from MASSEY RESEARCH ONLINE